Gallium3D Gets New Geometry Shader Support

[phoronix]

Phoronix: Gallium3D Gets New Geometry Shader Support

Zack Rusin has just shared with the Mesa development community a plethora of new code. This is not code for yet another state tracker or new hardware driver, but rather it brings new geometry shader support to the Gallium3D infrastructure...

http://www.phoronix.com/vr.php?view=Nzg0MQ

[V!NCENT]

Wasn't the geometry shading what made Direct3D version 10 so awesome and wasn´t it what made people so angry about the lack of it in OpenGL 3.0?

[Remco]

Wasn't the geometry shading what made Direct3D version 10 so awesome and wasn´t it what made people so angry about the lack of it in OpenGL 3.0?

I don't think so. Geometry shaders have been available as an OpenGL extension long before DirectX 10 was out. What made people angry about OpenGL 3.0 was that it didn't completely throw out the old standard and replace it with a shiny new object oriented one.

[Louise]

I wonder if something like this can be used for a cool Compiz effect?

[Remco]

I wonder if something like this can be used for a cool Compiz effect?

Some kickass tessellation of the desktop cube, maybe? :P

[DeepDayze]

Or why not add some fancy textures to the cube such as granite frames to each side

[Svartalf]

Wasn't the geometry shading what made Direct3D version 10 so awesome and wasn´t it what made people so angry about the lack of it in OpenGL 3.0?

Yes to the first question, no to the second.

There's extensions that support that functionality that were being ratified by the ARB during the period that everyone was beta-testing those D3D10 drivers. Plans were afoot with at least one vendor to support the same ASAP under OpenGL.

What everyone was so angry about wasn't the lack of Geometry shaders, it was the keeping of legacy (i.e. immediate mode rendering and things like it) interfaces and the omission of the new object model and API edge. 3.0 isn't the revolutionary change they'd originally promised for it- it's very weakly evolutionary in many ways.

[V!NCENT]

Thank you, Remco and Svartalf for the insight.

From what I understood is that OpenGl 3.0 was supposed to be completely new, that much I know, and it should also adapt to Direct3D 10ish hardware, right? From what I remember, but correct me if I am wrong, is that with the start of Direct3D 10 there is no longer a graphical pipeline, is that correct? Which should mean that one could get more performance out of this new architecture.

I thought that geometry shaders made that possible, but apparently they don't?

[thefirstm]

Wasn't the geometry shading what made Direct3D version 10 so awesome and wasn´t it what made people so angry about the lack of it in OpenGL 3.0?

Direct3D? Awesome? I think not....

[BlackStar]

Direct3D? Awesome? I think not....

Well, it is, if we want to be honest with ourselves. OpenGL 3.2 manages to close the gap somewhat, but it's still far behind D3D10 as far as API design, ease of use and stability is concerned. (Yes, it offers more functionality in general, but (a) it's still missing binary shaders and (b) 50% of the market is using Intel IGPs, which are synonymous with "bad OpenGL support" and (c) our OSS drivers don't even support GL2.1, much less 3.2).